# IBM Research

At IBM Research, we're inventing what's next in AI, quantum computing, and hybrid cloud to shape the world ahead.

This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles.

## SC26: IBM shows progress in bringing quantum & HPC together

DevFeed: [SC26: IBM shows progress in bringing quantum & HPC together](<https://devfeed.tech/articles/sc26-ibm-shows-progress-in-bringing-quantum-hpc-together-136374.md>)

Original publisher: [Read original article](<https://research.ibm.com/blog/sc26>)

Author: Rafi Letzter

Published: 2026-11-16T16:00:00Z

Content type: news

Language: en

Sources: [IBM Research](<https://devfeed.tech/sources/ibm-research.md>)

Topics: [Quantum Computing](<https://devfeed.tech/topics/quantum-computing.md>), [High-Performance Computing](<https://devfeed.tech/topics/high-performance-computing.md>), [Supercomputing](<https://devfeed.tech/topics/supercomputing.md>), [scheduling](<https://devfeed.tech/topics/scheduling.md>), [Tool](<https://devfeed.tech/topics/tool.md>), [ibm](<https://devfeed.tech/topics/ibm.md>)

Tags: [2026](<https://devfeed.tech/tags/2026.md>), [accelerators](<https://devfeed.tech/tags/accelerators.md>), [amd](<https://devfeed.tech/tags/amd.md>), [co-design](<https://devfeed.tech/tags/co-design.md>), [conference](<https://devfeed.tech/tags/conference.md>), [decoder](<https://devfeed.tech/tags/decoder.md>), [error-correction](<https://devfeed.tech/tags/error-correction.md>), [hpc](<https://devfeed.tech/tags/hpc.md>), [ibm](<https://devfeed.tech/tags/ibm.md>), [news](<https://devfeed.tech/tags/news.md>), [partners](<https://devfeed.tech/tags/partners.md>), [quantum](<https://devfeed.tech/tags/quantum.md>), [quantum-error-correction-mitigation](<https://devfeed.tech/tags/quantum-error-correction-mitigation.md>), [quantum-systems](<https://devfeed.tech/tags/quantum-systems.md>), [sc26](<https://devfeed.tech/tags/sc26.md>), [tools](<https://devfeed.tech/tags/tools.md>), [workflows](<https://devfeed.tech/tags/workflows.md>)

### AI overview

IBM and partners are integrating quantum processors into established HPC workflows through scheduler tools, hybrid computing platforms, and reusable managed workflows. QRMI lets Slurm jobs request a QPU alongside CPUs and GPUs, while IBM and NVIDIA are integrating it with Slurm HRES. In research, an SQD calculation of an iron-sulfur cluster on RIKEN's ROQUO supercomputer came within 4.3 millihartree of DMRG, 79x closer than the team's earlier result; a separate quantum-HPC workflow simulated protein-ligand complexes up to 12,635 atoms. For error-correction research, IBM and AMD's open-source Pelagos uses Triton on AMD and NVIDIA GPUs to simulate decoder tests, running up to 97x faster than a single CPU core and scaling across multiple GPUs. IBM also describes managed quantum-classical workflows and options ranging from a free Open Plan to paid credits, Flex Plan access, or dedicated on-premises systems. The excerpt reports a planned IBM-NVIDIA-Fermilab co-design platform and does not provide its launch date.

### Source excerpt

At SC26 in Chicago, IBM and partners show how quantum computers are integrating into workflows and tools that the supercomputing workforce already runs.

## IBM, AMD advance quantum error correction decoding at scale

DevFeed: [IBM, AMD advance quantum error correction decoding at scale](<https://devfeed.tech/articles/ibm-amd-advance-quantum-error-correction-decoding-at-scale-136373.md>)

Original publisher: [Read original article](<https://research.ibm.com/blog/pelagos>)

Author: Blake Johnson; Robert Davis

Published: 2026-11-16T15:59:00Z

Content type: news

Language: en

Sources: [IBM Research](<https://devfeed.tech/sources/ibm-research.md>)

Topics: [Quantum Computing](<https://devfeed.tech/topics/quantum-computing.md>), [Decoding](<https://devfeed.tech/topics/decoding.md>), [rocm](<https://devfeed.tech/topics/rocm.md>), [ibm](<https://devfeed.tech/topics/ibm.md>)

Tags: [acceleration](<https://devfeed.tech/tags/acceleration.md>), [accelerators](<https://devfeed.tech/tags/accelerators.md>), [algorithms](<https://devfeed.tech/tags/algorithms.md>), [amd](<https://devfeed.tech/tags/amd.md>), [decoding](<https://devfeed.tech/tags/decoding.md>), [evaluation](<https://devfeed.tech/tags/evaluation.md>), [ibm](<https://devfeed.tech/tags/ibm.md>), [open-source](<https://devfeed.tech/tags/open-source.md>), [quantum](<https://devfeed.tech/tags/quantum.md>), [quantum-error-correction](<https://devfeed.tech/tags/quantum-error-correction.md>), [quantum-research](<https://devfeed.tech/tags/quantum-research.md>), [quantum-software](<https://devfeed.tech/tags/quantum-software.md>), [release](<https://devfeed.tech/tags/release.md>), [scale](<https://devfeed.tech/tags/scale.md>)

### AI overview

IBM and AMD released Pelagos, an open-source framework for evaluating quantum error-correction decoders at scale. Built with Triton, it supports AMD and NVIDIA GPUs and combines GPU acceleration with a reusable software architecture. Multi-GPU work uses process-level orchestration, parallel host-side operations and shared memory to reduce coordination overhead. Tests on AMD Instinct MI300X and MI355X showed up to roughly 7.5x speedup in the GPU decoding stage across eight GPUs versus one; the team also deployed Pelagos on ORNL's Lux HPC system. The excerpt says earlier studies found up to 97x higher throughput than a single CPU core. An accompanying survey reviews decoding methods and platforms including CPUs, GPUs, FPGAs and specialized accelerators, and argues that classical infrastructure and hardware-software co-design are key parts of building fault-tolerant quantum systems.

### Source excerpt

New open-source tool from IBM and AMD helps researchers evaluate quantum error-correction decoders at scale for future fault-tolerant quantum systems.

## What is a spin qubit?

DevFeed: [What is a spin qubit?](<https://devfeed.tech/articles/what-is-a-spin-qubit-130672.md>)

Original publisher: [Read original article](<https://research.ibm.com/blog/spin-qubits-explained>)

Author: Ryan Mandelbaum

Published: 2026-11-12T15:30:00Z

Content type: article

Language: en

Sources: [IBM Research](<https://devfeed.tech/sources/ibm-research.md>)

Topics: [Quantum Computing](<https://devfeed.tech/topics/quantum-computing.md>), [quantum mechanics](<https://devfeed.tech/topics/quantum-mechanics.md>)

Tags: [2](<https://devfeed.tech/tags/2.md>), [about](<https://devfeed.tech/tags/about.md>), [advanced](<https://devfeed.tech/tags/advanced.md>), [algorithms](<https://devfeed.tech/tags/algorithms.md>), [also](<https://devfeed.tech/tags/also.md>), [analogue](<https://devfeed.tech/tags/analogue.md>), [anderon](<https://devfeed.tech/tags/anderon.md>), [apply](<https://devfeed.tech/tags/apply.md>), [applying](<https://devfeed.tech/tags/applying.md>), [approach](<https://devfeed.tech/tags/approach.md>), [building](<https://devfeed.tech/tags/building.md>), [computers](<https://devfeed.tech/tags/computers.md>), [encoding](<https://devfeed.tech/tags/encoding.md>), [explainer](<https://devfeed.tech/tags/explainer.md>), [information](<https://devfeed.tech/tags/information.md>), [possible](<https://devfeed.tech/tags/possible.md>), [quantum](<https://devfeed.tech/tags/quantum.md>), [quantum-computers](<https://devfeed.tech/tags/quantum-computers.md>), [quantum-research](<https://devfeed.tech/tags/quantum-research.md>), [qubit](<https://devfeed.tech/tags/qubit.md>), [spin](<https://devfeed.tech/tags/spin.md>), [spin-qubits](<https://devfeed.tech/tags/spin-qubits.md>), [what-is](<https://devfeed.tech/tags/what-is.md>)

### AI overview

Spin qubits encode quantum information in electron spin states confined in quantum dots. In the Loss-DiVincenzo design, one electron per dot is controlled with microwave pulses, and voltage pulses couple neighboring electrons through exchange interactions. Exchange-only designs use three dots and electrons per qubit and rely on voltage pulses for control. HRL reports a 54-dot system configurable for up to 18 qubits that ran a simple quantum-error-detecting code, controlled by a custom CMOS chip at 4 K. The text presents spin qubits as fast, high-fidelity candidates compatible with semiconductor fabrication; their simpler voltage-based control may reduce room-temperature wiring. They still lag superconducting processors in measurement speed and current qubit count, and scaling requires more uniform qubits, better fabrication yield, lower controller power, and less wiring. The source expects superconducting qubits to reach fault tolerance first.

### Source excerpt

A possible path toward building larger quantum computers involves encoding quantum information in the spins of electrons.

## Tracking progress on the path to IBM Quantum Starling

DevFeed: [Tracking progress on the path to IBM Quantum Starling](<https://devfeed.tech/articles/tracking-progress-on-the-path-to-ibm-quantum-starling-129812.md>)

Original publisher: [Read original article](<https://research.ibm.com/blog/path-to-starling>)

Author: Matthias Steffen; Matthew Marwick

Published: 2026-11-11T13:00:00Z

Content type: article

Language: en

Sources: [IBM Research](<https://devfeed.tech/sources/ibm-research.md>)

Topics: [Quantum Computing](<https://devfeed.tech/topics/quantum-computing.md>), [quantum mechanics](<https://devfeed.tech/topics/quantum-mechanics.md>), [Distributed Systems & Parallel Computing](<https://devfeed.tech/topics/distributed-systems-parallel-computing.md>), [High-Performance Computing](<https://devfeed.tech/topics/high-performance-computing.md>), [ibm](<https://devfeed.tech/topics/ibm.md>)

Tags: [computing](<https://devfeed.tech/tags/computing.md>), [fault-tolerant-quantum-computing](<https://devfeed.tech/tags/fault-tolerant-quantum-computing.md>), [ibm](<https://devfeed.tech/tags/ibm.md>), [ibm-quantum](<https://devfeed.tech/tags/ibm-quantum.md>), [parallel](<https://devfeed.tech/tags/parallel.md>), [progress](<https://devfeed.tech/tags/progress.md>), [quantum](<https://devfeed.tech/tags/quantum.md>), [quantum-architectures](<https://devfeed.tech/tags/quantum-architectures.md>), [quantum-computing](<https://devfeed.tech/tags/quantum-computing.md>), [quantum-error-correction-mitigation](<https://devfeed.tech/tags/quantum-error-correction-mitigation.md>), [research](<https://devfeed.tech/tags/research.md>), [starling](<https://devfeed.tech/tags/starling.md>), [technology](<https://devfeed.tech/tags/technology.md>), [tracking](<https://devfeed.tech/tags/tracking.md>)

### AI overview

IBM reports progress toward its planned 2029 fault-tolerant Quantum Starling, designed for circuits with 100 million operations on 200 qubits. Its modular qLDPC architecture is estimated to need about 10 times fewer physical qubits than prior approaches at the same logical error rate. Updates include validated c-couplers for dense on-chip connections, code and coupler changes that cut required Josephson junctions from about 5,000 to 2,000, and a decoder whose observed performance meets Starling requirements. Two Nighthawk processors have been linked with l-couplers at an average 1% error rate, while Starling will require over 100 such links per module at 0.1%, making reliability and repeatability ongoing challenges. IBM has also tested modular cryostats and cryogenic control ASICs that match or exceed room-temperature gate performance. The overall architecture is unchanged, but Starling remains a few years away, with further work including refining logical processing units and reducing magic-state injection overhead.

### Source excerpt

Advances across parallel technology areas are laying the groundwork for fault-tolerant quantum computing by the end of the decade.

## QDC 2026: what quantum developers need to know

DevFeed: [QDC 2026: what quantum developers need to know](<https://devfeed.tech/articles/qdc-2026-what-quantum-developers-need-to-know-129810.md>)

Original publisher: [Read original article](<https://research.ibm.com/blog/dev-tools-qdc-2026>)

Author: Genya Crossman; Sanskriti Deva; Robert Davis

Published: 2026-11-11T11:59:00Z

Content type: article

Language: en

Sources: [IBM Research](<https://devfeed.tech/sources/ibm-research.md>)

Topics: [Quantum Computing](<https://devfeed.tech/topics/quantum-computing.md>), [High-Performance Computing](<https://devfeed.tech/topics/high-performance-computing.md>), [Quantum Enablement](<https://devfeed.tech/topics/quantum-enablement.md>), [Tool](<https://devfeed.tech/topics/tool.md>), [Learning](<https://devfeed.tech/topics/learning.md>)

Tags: [1-5](<https://devfeed.tech/tags/1-5.md>), [2](<https://devfeed.tech/tags/2.md>), [2026](<https://devfeed.tech/tags/2026.md>), [3](<https://devfeed.tech/tags/3.md>), [about](<https://devfeed.tech/tags/about.md>), [abstractions](<https://devfeed.tech/tags/abstractions.md>), [academic](<https://devfeed.tech/tags/academic.md>), [academic-research](<https://devfeed.tech/tags/academic-research.md>), [conference](<https://devfeed.tech/tags/conference.md>), [developer](<https://devfeed.tech/tags/developer.md>), [developers](<https://devfeed.tech/tags/developers.md>), [guide](<https://devfeed.tech/tags/guide.md>), [hardware](<https://devfeed.tech/tags/hardware.md>), [ibm](<https://devfeed.tech/tags/ibm.md>), [ibm-quantum](<https://devfeed.tech/tags/ibm-quantum.md>), [know](<https://devfeed.tech/tags/know.md>), [need](<https://devfeed.tech/tags/need.md>), [news](<https://devfeed.tech/tags/news.md>), [quantum](<https://devfeed.tech/tags/quantum.md>), [quantum-enablement](<https://devfeed.tech/tags/quantum-enablement.md>), [quantum-error-correction-mitigation](<https://devfeed.tech/tags/quantum-error-correction-mitigation.md>), [quantum-software](<https://devfeed.tech/tags/quantum-software.md>), [quantum-systems](<https://devfeed.tech/tags/quantum-systems.md>), [software](<https://devfeed.tech/tags/software.md>), [workflows](<https://devfeed.tech/tags/workflows.md>)

### AI overview

QDC 2026's main developer takeaway is that IBM is exposing more control over real hardware while adding tools to build, inspect, and scale quantum workflows. Nighthawk r2 offers 120 programmable qubits, pulse-level access, dynamic circuits and fast reset; the excerpt also describes a repetition-code experiment with one million measurement and reset rounds, though `for_loop` and the `qec_experimental` calibration are experimental, and `for_loop` is currently limited to Executor and client-side Sampler jobs. The directed execution model pairs Samplomatic with Executor for inspectable circuit transformations and execution; client-side Sampler and Estimator expose pre- and post-processing, including error mitigation, for customization. Managed quantum-classical workflows require Premium or Flex plans, while dry-run usage estimates consume no QPU time. For applications, Qiskit Fermions and Bosons defer encoding until transpilation, and `ffsim` supports classical validation before hardware runs. QRMI offers a vendor-agnostic interface for integrating supported quantum backends into HPC workflows. The excerpt also outlines expanded free access for qualifying active Open Plan users.

### Source excerpt

A developer's guide to the hardware, software, and workflows unveiled at IBM Quantum Developer Conference 2026.

## Building policy enforcement into the AI stack with Praxis

DevFeed: [Building policy enforcement into the AI stack with Praxis](<https://devfeed.tech/articles/building-policy-enforcement-into-the-ai-stack-with-praxis-124792.md>)

Original publisher: [Read original article](<https://research.ibm.com/blog/building-policy-enforcement-into-the-ai-stack-with-praxis>)

Author: Malgorzata Steinder; Frederico Araujo; Ian Molloy

Published: 2026-11-09T15:30:00Z

Content type: article

Language: en

Sources: [IBM Research](<https://devfeed.tech/sources/ibm-research.md>)

Topics: [LLM security](<https://devfeed.tech/topics/llm-security.md>), [AI Infrastructure](<https://devfeed.tech/topics/ai-infrastructure.md>), [policy-as-code](<https://devfeed.tech/topics/policy-as-code.md>)

Tags: [access-control](<https://devfeed.tech/tags/access-control.md>), [accounting](<https://devfeed.tech/tags/accounting.md>), [agent](<https://devfeed.tech/tags/agent.md>), [agent-authorization](<https://devfeed.tech/tags/agent-authorization.md>), [agent-identity](<https://devfeed.tech/tags/agent-identity.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-gateway](<https://devfeed.tech/tags/ai-gateway.md>), [ai-platform](<https://devfeed.tech/tags/ai-platform.md>), [ai-security](<https://devfeed.tech/tags/ai-security.md>), [ai-security-research](<https://devfeed.tech/tags/ai-security-research.md>), [also](<https://devfeed.tech/tags/also.md>), [building](<https://devfeed.tech/tags/building.md>), [data-and-ai-security](<https://devfeed.tech/tags/data-and-ai-security.md>), [engine](<https://devfeed.tech/tags/engine.md>), [gateways](<https://devfeed.tech/tags/gateways.md>), [inference](<https://devfeed.tech/tags/inference.md>), [policy](<https://devfeed.tech/tags/policy.md>), [policy-enforcement](<https://devfeed.tech/tags/policy-enforcement.md>), [security](<https://devfeed.tech/tags/security.md>), [technical-note](<https://devfeed.tech/tags/technical-note.md>), [trustworthy-ai](<https://devfeed.tech/tags/trustworthy-ai.md>), [with](<https://devfeed.tech/tags/with.md>)

### AI overview

Praxis embeds IBM and Red Hat's open-source Praxis Policy Engine (PPE) as a policy filter for AI and ordinary traffic. PPE makes tools, prompts, resources, model calls and other agents explicit policy targets, combining identity, delegation, content controls and audit in a reference monitor based on verified state outside the model's control. Teams can specify access rules and behavior policies and deploy enforcement at a gateway, beside an agent or inside a framework; each placement covers only operations that cross it. PPE supports application-layer enforcement, including a laptop policy build in Rosso Cortex, and integrates with ContextForge and Mellea. It can emit OCSF audit records for policy decisions, with optional signatures and hash chaining. IBM describes ongoing work on combining gateway enforcement with host isolation and testing bypass resistance, usability, portability and assurance across agent tool calls and delegation.

### Source excerpt

From inference through gateways, agents, and policy controls, the Praxis Policy Engine is a programmable security guard for AI agents.

## Chicago's bid to lead the country in quantum computing

DevFeed: [Chicago's bid to lead the country in quantum computing](<https://devfeed.tech/articles/chicago-s-bid-to-lead-the-country-in-quantum-computing-121802.md>)

Original publisher: [Read original article](<https://research.ibm.com/blog/chicago-quantum-computer>)

Author: Rafi Letzter

Published: 2026-11-06T05:00:00Z

Content type: news

Language: en

Sources: [IBM Research](<https://devfeed.tech/sources/ibm-research.md>)

Topics: [Quantum Computing](<https://devfeed.tech/topics/quantum-computing.md>), [High-Performance Computing](<https://devfeed.tech/topics/high-performance-computing.md>)

Tags: [algorithm](<https://devfeed.tech/tags/algorithm.md>), [algorithms](<https://devfeed.tech/tags/algorithms.md>), [applications](<https://devfeed.tech/tags/applications.md>), [chicago-quantum-exchange](<https://devfeed.tech/tags/chicago-quantum-exchange.md>), [collaboration](<https://devfeed.tech/tags/collaboration.md>), [collective](<https://devfeed.tech/tags/collective.md>), [company](<https://devfeed.tech/tags/company.md>), [computer](<https://devfeed.tech/tags/computer.md>), [computing](<https://devfeed.tech/tags/computing.md>), [convergence](<https://devfeed.tech/tags/convergence.md>), [discovery](<https://devfeed.tech/tags/discovery.md>), [drug-discovery](<https://devfeed.tech/tags/drug-discovery.md>), [economic-development](<https://devfeed.tech/tags/economic-development.md>), [ecosystem](<https://devfeed.tech/tags/ecosystem.md>), [ibm](<https://devfeed.tech/tags/ibm.md>), [industry](<https://devfeed.tech/tags/industry.md>), [news](<https://devfeed.tech/tags/news.md>), [quantum](<https://devfeed.tech/tags/quantum.md>), [quantum-community](<https://devfeed.tech/tags/quantum-community.md>), [quantum-computing](<https://devfeed.tech/tags/quantum-computing.md>), [quantum-network](<https://devfeed.tech/tags/quantum-network.md>), [quantum-systems](<https://devfeed.tech/tags/quantum-systems.md>)

### AI overview

Illinois is building a quantum computing hub around the National Quantum Algorithm Center and the future Illinois Quantum and Microelectronics Park, backed by a $500 million state investment and partners including IBM, the University of Chicago, and the University of Illinois. IBM has installed a Quantum System Two with a Nighthawk r2 processor at Hyde Park Labs; it will later move to the park, which is intended to house the center and computer. The excerpt describes regional research and application efforts, including IBM and UChicago's LASSQD method, which uses quantum hardware for molecular fragments and modeled a 26-electron, 24-orbital iron porphyrin fragment, and projects exploring drug discovery and power-grid optimization. Researchers and organizers present the facility as both a practical resource and a public symbol, while emphasizing decades of Midwest talent and institution-building as the basis for Illinois' ambitions.

### Source excerpt

New IBM Quantum System Two supports Illinois' place at the center of the American quantum computing industry.

## ORNL, IBM, AMD launch co-design platform for QEC

DevFeed: [ORNL, IBM, AMD launch co-design platform for QEC](<https://devfeed.tech/articles/ornl-ibm-amd-launch-co-design-platform-for-qec-119360.md>)

Original publisher: [Read original article](<https://research.ibm.com/blog/ornl-amd-error-correction-co-design-platform>)

Author: Rafi Letzter

Published: 2026-11-05T13:00:00Z

Content type: article

Language: en

Sources: [IBM Research](<https://devfeed.tech/sources/ibm-research.md>)

Topics: [High-Performance Computing](<https://devfeed.tech/topics/high-performance-computing.md>), [ibm](<https://devfeed.tech/topics/ibm.md>)

Tags: [amd](<https://devfeed.tech/tags/amd.md>), [co-design](<https://devfeed.tech/tags/co-design.md>), [hardware](<https://devfeed.tech/tags/hardware.md>), [hpc](<https://devfeed.tech/tags/hpc.md>), [ibm](<https://devfeed.tech/tags/ibm.md>), [launch](<https://devfeed.tech/tags/launch.md>), [news](<https://devfeed.tech/tags/news.md>), [oak-ridge](<https://devfeed.tech/tags/oak-ridge.md>), [ornl](<https://devfeed.tech/tags/ornl.md>), [platform](<https://devfeed.tech/tags/platform.md>), [quantum](<https://devfeed.tech/tags/quantum.md>), [quantum-community](<https://devfeed.tech/tags/quantum-community.md>), [quantum-network](<https://devfeed.tech/tags/quantum-network.md>), [science](<https://devfeed.tech/tags/science.md>)

### AI overview

ORNL, IBM, and AMD have launched a quantum error correction co-design platform at IBM's site in Poughkeepsie, pairing IBM Quantum Nighthawk r2 processors with ORNL's Umbra HPC system, powered by AMD MI355X GPUs and 5th Gen EPYC CPUs. Co-locating quantum and classical hardware is intended to reduce delays in decoding and measurement-driven workflows. Researchers can develop and validate algorithms and error correction methods on real quantum systems without building the infrastructure themselves; the platform also enables exploration of GPU-based low-latency decoding and hierarchical codes. The text describes a research environment, not a completed fault-tolerant computer, and says its work will inform IBM Quantum Starling, planned for 2029.

### Source excerpt

The U.S. DOE's Quantum Science Center at Oak Ridge National Lab is leading the integration of IBM quantum hardware with ORNL HPC clusters, powered by AMD.

## How MEMS inertial sensors can rival GPS

DevFeed: [How MEMS inertial sensors can rival GPS](<https://devfeed.tech/articles/how-mems-inertial-sensors-can-rival-gps-119359.md>)

Original publisher: [Read original article](<https://research.ibm.com/blog/how-mems-inertial-sensors-can-rival-gps>)

Author: Matthew Marwick

Published: 2026-11-05T13:00:00Z

Content type: article

Language: en

Sources: [IBM Research](<https://devfeed.tech/sources/ibm-research.md>)

Topics: [navigation](<https://devfeed.tech/topics/navigation.md>), [Physical AI](<https://devfeed.tech/topics/physical-ai.md>)

Tags: [about](<https://devfeed.tech/tags/about.md>), [acceleration](<https://devfeed.tech/tags/acceleration.md>), [accelerometers](<https://devfeed.tech/tags/accelerometers.md>), [account](<https://devfeed.tech/tags/account.md>), [added](<https://devfeed.tech/tags/added.md>), [addition](<https://devfeed.tech/tags/addition.md>), [advanced](<https://devfeed.tech/tags/advanced.md>), [advanced-packaging](<https://devfeed.tech/tags/advanced-packaging.md>), [aerospace](<https://devfeed.tech/tags/aerospace.md>), [algorithm](<https://devfeed.tech/tags/algorithm.md>), [algorithms](<https://devfeed.tech/tags/algorithms.md>), [also](<https://devfeed.tech/tags/also.md>), [angular](<https://devfeed.tech/tags/angular.md>), [application](<https://devfeed.tech/tags/application.md>), [breakthrough](<https://devfeed.tech/tags/breakthrough.md>), [can](<https://devfeed.tech/tags/can.md>), [explainer](<https://devfeed.tech/tags/explainer.md>), [form](<https://devfeed.tech/tags/form.md>), [gps](<https://devfeed.tech/tags/gps.md>), [hardware-technology](<https://devfeed.tech/tags/hardware-technology.md>), [mems-inertial-sensing](<https://devfeed.tech/tags/mems-inertial-sensing.md>), [navigation](<https://devfeed.tech/tags/navigation.md>), [sensing](<https://devfeed.tech/tags/sensing.md>), [sensors](<https://devfeed.tech/tags/sensors.md>), [technology](<https://devfeed.tech/tags/technology.md>)

### AI overview

HRL Laboratories' MEMS-based IMU uses accelerometers and gyroscopes to estimate movement and position through dead reckoning when GPS signals are unavailable. Its AutoCal system takes simultaneous sensor measurements in opposite directions to isolate and remove bias, addressing drift that the text says has limited typical MEMS IMUs to only minutes of navigation. The device is about the size of a soup can, with custom ASICs, compact packaging, and wafer-based manufacturing intended to reduce size and cost. HRL aims to shrink it to hockey-puck size; production is expected to begin in 2027. The text describes an intended GPS alternative, but gives no measured accuracy results.

### Source excerpt

A breakthrough in self-calibration technology is revitalizing an old form of navigation.

## How Zurich is shaping the future of algorithms, 70 years in

DevFeed: [How Zurich is shaping the future of algorithms, 70 years in](<https://devfeed.tech/articles/how-zurich-is-shaping-the-future-of-algorithms-70-years-in-106374.md>)

Original publisher: [Read original article](<https://research.ibm.com/blog/celebrating-70-years-ibm-research-europe>)

Author: Mariana Rajado Silva

Published: 2026-10-29T14:15:00Z

Content type: news

Language: en

Sources: [IBM Research](<https://devfeed.tech/sources/ibm-research.md>)

Topics: [Algorithms](<https://devfeed.tech/topics/algorithms.md>), [Cryptography](<https://devfeed.tech/topics/cryptography.md>), [AI Models](<https://devfeed.tech/topics/ai-models.md>), [quantum mechanics](<https://devfeed.tech/topics/quantum-mechanics.md>)

Tags: [agentic](<https://devfeed.tech/tags/agentic.md>), [agentic-ai](<https://devfeed.tech/tags/agentic-ai.md>), [ai](<https://devfeed.tech/tags/ai.md>), [algorithms](<https://devfeed.tech/tags/algorithms.md>), [applications](<https://devfeed.tech/tags/applications.md>), [computer-science](<https://devfeed.tech/tags/computer-science.md>), [computing](<https://devfeed.tech/tags/computing.md>), [cpus](<https://devfeed.tech/tags/cpus.md>), [cryptography](<https://devfeed.tech/tags/cryptography.md>), [digital-twins](<https://devfeed.tech/tags/digital-twins.md>), [future](<https://devfeed.tech/tags/future.md>), [ibm](<https://devfeed.tech/tags/ibm.md>), [lab](<https://devfeed.tech/tags/lab.md>), [q-a](<https://devfeed.tech/tags/q-a.md>), [vision](<https://devfeed.tech/tags/vision.md>)

### AI overview

Alessandro Curioni says IBM Research is developing foundational algorithms that combine classical computing, quantum computing and AI across five areas: differential equations and dynamical systems; linear algebra and Hamiltonian problems; optimization and combinatorial problems; probability, statistics and stochastic processes; and cryptography. He envisions CPUs, GPUs and QPUs working together under an agentic AI orchestration layer, enabling new approaches to complex problems and more accurate digital twins for science and business. He also credits IBM Research's ability to attract and retain talent, its global reach, and its connection between research and industry as key strengths. The excerpt presents his views and priorities, not specific algorithm results or implementation requirements.

### Source excerpt

As IBM's Zurich Lab turns 70, director Alessandro Curioni speaks about the celebrations, his work on Algorithms & Applications, and his vision for the future.

## Modular quantum computing takes shape with l-couplers

DevFeed: [Modular quantum computing takes shape with l-couplers](<https://devfeed.tech/articles/modular-quantum-computing-takes-shape-with-l-couplers-102057.md>)

Original publisher: [Read original article](<https://research.ibm.com/blog/connecting-quantum-processors-with-l-couplers>)

Author: Ryan Mandelbaum; Neereja Sundaresan

Published: 2026-10-27T12:00:00Z

Content type: article

Language: en

Sources: [IBM Research](<https://devfeed.tech/sources/ibm-research.md>)

Topics: [Quantum Computing](<https://devfeed.tech/topics/quantum-computing.md>), [ibm](<https://devfeed.tech/topics/ibm.md>)

Tags: [algorithms](<https://devfeed.tech/tags/algorithms.md>), [architecture](<https://devfeed.tech/tags/architecture.md>), [computing](<https://devfeed.tech/tags/computing.md>), [error-correction](<https://devfeed.tech/tags/error-correction.md>), [fault-tolerance](<https://devfeed.tech/tags/fault-tolerance.md>), [fault-tolerant-quantum-computing](<https://devfeed.tech/tags/fault-tolerant-quantum-computing.md>), [hardware](<https://devfeed.tech/tags/hardware.md>), [ibm](<https://devfeed.tech/tags/ibm.md>), [information](<https://devfeed.tech/tags/information.md>), [modular](<https://devfeed.tech/tags/modular.md>), [news](<https://devfeed.tech/tags/news.md>), [quantum](<https://devfeed.tech/tags/quantum.md>), [quantum-computers](<https://devfeed.tech/tags/quantum-computers.md>), [quantum-computing](<https://devfeed.tech/tags/quantum-computing.md>), [quantum-systems](<https://devfeed.tech/tags/quantum-systems.md>), [scaling](<https://devfeed.tech/tags/scaling.md>)

### AI overview

IBM researchers connected two Quantum Nighthawk R1 processors inside a dilution refrigerator using four 80-centimeter l-couplers. Quantum state transfer reached up to 99.3% fidelity, averaging 99%; distant CZ gates averaged 94.75% fidelity across the four links and enabled entangled states shared between processors. The links include coaxial cables, chip components, l-qubits, tunable couplers and controllers, and are installed after the processors are mounted in the refrigerator. The design is presented as compatible with Nighthawk and scalable toward Starling, addressing the need for multiple inter-chip connections in modular systems. The excerpt describes this as a scaling milestone, while noting that yield, reproducibility, error rates and the number of couplers still need improvement.

### Source excerpt

IBM researchers achieve a critical advance for scaling quantum computers.

## IBM and Kaggle launch Java framework migration leaderboard

DevFeed: [IBM and Kaggle launch Java framework migration leaderboard](<https://devfeed.tech/articles/ibm-and-kaggle-launch-java-framework-migration-leaderboard-102058.md>)

Original publisher: [Read original article](<https://research.ibm.com/blog/ibm-and-kaggle-launch-leaderboard-for-java-framework-migration>)

Author: Bridget McGinn; Raju Pavuluri; Advait Pavuluri; Rahul Krishna; Srikanth Tamilselvam; Ashita Saxena; George Safta; Michele Merler

Published: 2026-10-27T12:00:00Z

Content type: news

Language: en

Sources: [IBM Research](<https://devfeed.tech/sources/ibm-research.md>)

Topics: [Agent Harness](<https://devfeed.tech/topics/agent-harness.md>), [Java](<https://devfeed.tech/topics/java.md>), [Jakarta EE](<https://devfeed.tech/topics/jakarta-ee.md>), [Quarkus](<https://devfeed.tech/topics/quarkus.md>), [Spring Framework](<https://devfeed.tech/topics/spring-framework.md>), [Kaggle](<https://devfeed.tech/topics/kaggle.md>), [ibm](<https://devfeed.tech/topics/ibm.md>)

Tags: [agent](<https://devfeed.tech/tags/agent.md>), [agent-harness](<https://devfeed.tech/tags/agent-harness.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-for-code](<https://devfeed.tech/tags/ai-for-code.md>), [ai-for-it](<https://devfeed.tech/tags/ai-for-it.md>), [enterprise](<https://devfeed.tech/tags/enterprise.md>), [framework](<https://devfeed.tech/tags/framework.md>), [harness](<https://devfeed.tech/tags/harness.md>), [ibm](<https://devfeed.tech/tags/ibm.md>), [java](<https://devfeed.tech/tags/java.md>), [java-migration](<https://devfeed.tech/tags/java-migration.md>), [kaggle](<https://devfeed.tech/tags/kaggle.md>), [launch](<https://devfeed.tech/tags/launch.md>), [leaderboard](<https://devfeed.tech/tags/leaderboard.md>), [migration](<https://devfeed.tech/tags/migration.md>), [model](<https://devfeed.tech/tags/model.md>), [technical-note](<https://devfeed.tech/tags/technical-note.md>), [validation](<https://devfeed.tech/tags/validation.md>)

### AI overview

ScarfBench tests whether coding-agent configurations can migrate enterprise Java applications across Jakarta EE, Quarkus, and Spring while preserving behavior, checking builds, deployment, and behavioral tests. In 204 tasks, Gemini 4 Argon with Antigravity led the smoke-test leaderboard at 75.2%, while Claude Opus 5.5 with Claude Code had the highest all-tests-pass rate at 72.1%; results measure each model paired with its harness, not the model alone. Compilation was much less discriminating than behavioral fidelity: about 72% of 745 classified failures occurred at runtime, including assertion mismatches and lost routes. IBM and Kaggle say Kaggle independently reproduced the evaluation, sponsored compute, and will maintain the benchmark and add models over time. Public submissions were not available at launch but were on the roadmap.

### Source excerpt

ScarfBench evaluates model-and-agent-harness combinations for enterprise Java migration, pairing migration validation with reproduced runs and a leaderboard.

## PyTorch opens the AI ecosystem to a new accelerator class

DevFeed: [PyTorch opens the AI ecosystem to a new accelerator class](<https://devfeed.tech/articles/pytorch-opens-the-ai-ecosystem-to-a-new-accelerator-class-100549.md>)

Original publisher: [Read original article](<https://research.ibm.com/blog/pytorch-spyre-accelerator-software-ecosystem-updates>)

Author: Peter Hess

Published: 2026-10-26T12:00:00Z

Content type: article

Language: en

Sources: [IBM Research](<https://devfeed.tech/sources/ibm-research.md>)

Topics: [PyTorch](<https://devfeed.tech/topics/pytorch.md>), [MLOps](<https://devfeed.tech/topics/mlops.md>), [CI/CD](<https://devfeed.tech/topics/cicd.md>), [Distributed Systems & Parallel Computing](<https://devfeed.tech/topics/distributed-systems-parallel-computing.md>), [Developer experience](<https://devfeed.tech/topics/developer-experience.md>)

Tags: [accelerators](<https://devfeed.tech/tags/accelerators.md>), [adapters](<https://devfeed.tech/tags/adapters.md>), [adoption](<https://devfeed.tech/tags/adoption.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-accelerator](<https://devfeed.tech/tags/ai-accelerator.md>), [ai-hardware](<https://devfeed.tech/tags/ai-hardware.md>), [backend](<https://devfeed.tech/tags/backend.md>), [ci](<https://devfeed.tech/tags/ci.md>), [ci-cd](<https://devfeed.tech/tags/ci-cd.md>), [compatibility](<https://devfeed.tech/tags/compatibility.md>), [compiler](<https://devfeed.tech/tags/compiler.md>), [continuous-integration](<https://devfeed.tech/tags/continuous-integration.md>), [cpu](<https://devfeed.tech/tags/cpu.md>), [developers](<https://devfeed.tech/tags/developers.md>), [devices](<https://devfeed.tech/tags/devices.md>), [distributed-inference](<https://devfeed.tech/tags/distributed-inference.md>), [ecosystem](<https://devfeed.tech/tags/ecosystem.md>), [first](<https://devfeed.tech/tags/first.md>), [generative-ai](<https://devfeed.tech/tags/generative-ai.md>), [hybrid-cloud](<https://devfeed.tech/tags/hybrid-cloud.md>), [ibm](<https://devfeed.tech/tags/ibm.md>), [news](<https://devfeed.tech/tags/news.md>), [open-source](<https://devfeed.tech/tags/open-source.md>), [pytorch](<https://devfeed.tech/tags/pytorch.md>), [workflows](<https://devfeed.tech/tags/workflows.md>)

### AI overview

IBM's Spyre accelerator can run familiar PyTorch workflows through torch-spyre, which compiles PyTorch models for Spyre and uses adapters when operations need an equivalent compilable form. Spyre's dataflow execution differs from GPU instruction-stream execution; PyTorch and vLLM extension points provide a shared route to support it without rebuilding the larger software stack. The Spyre vLLM plugin has been validated with Gemma 4, Qwen3, and Granite 4.2, and its PyTorch backend also enables access through llm-d. A demonstration reports 1.7x faster prefill and 2.4x faster decode after moving graph construction off both data transfer and compute dispatch, for Granite 3.3 8B at batch size 1 and sequence length 1024. End-to-end model tests and continuous integration help catch compatibility issues that isolated unit tests may miss. Portability remains an active effort: on H100 GPUs, vLLM's hardware-agnostic layers delivered token throughput within 3.4% of native implementations across three models.

### Source excerpt

IBM has made Spyre a first-class PyTorch backend, giving developers the same workflows across devices.

## IBM supports award-winning projects in NEDO Challenge

DevFeed: [IBM supports award-winning projects in NEDO Challenge](<https://devfeed.tech/articles/ibm-supports-award-winning-projects-in-nedo-challenge-98478.md>)

Original publisher: [Read original article](<https://research.ibm.com/blog/ibm-nedo-challenge-2026>)

Author: Chris Nay; Toru Imai (今井 亨); Yuri Kobayashi

Published: 2026-10-23T04:00:00Z

Content type: news

Language: en

Sources: [IBM Research](<https://devfeed.tech/sources/ibm-research.md>)

Topics: [ibm](<https://devfeed.tech/topics/ibm.md>)

Tags: [algorithm](<https://devfeed.tech/tags/algorithm.md>), [all](<https://devfeed.tech/tags/all.md>), [analysis](<https://devfeed.tech/tags/analysis.md>), [animation](<https://devfeed.tech/tags/animation.md>), [applications](<https://devfeed.tech/tags/applications.md>), [approach](<https://devfeed.tech/tags/approach.md>), [arizona-state-university](<https://devfeed.tech/tags/arizona-state-university.md>), [award](<https://devfeed.tech/tags/award.md>), [awards](<https://devfeed.tech/tags/awards.md>), [business](<https://devfeed.tech/tags/business.md>), [challenge](<https://devfeed.tech/tags/challenge.md>), [competition](<https://devfeed.tech/tags/competition.md>), [computers](<https://devfeed.tech/tags/computers.md>), [ibm](<https://devfeed.tech/tags/ibm.md>), [japan](<https://devfeed.tech/tags/japan.md>), [news](<https://devfeed.tech/tags/news.md>), [quantum](<https://devfeed.tech/tags/quantum.md>), [quantum-community](<https://devfeed.tech/tags/quantum-community.md>), [quantum-computers](<https://devfeed.tech/tags/quantum-computers.md>), [quantum-network](<https://devfeed.tech/tags/quantum-network.md>), [real-world](<https://devfeed.tech/tags/real-world.md>), [research](<https://devfeed.tech/tags/research.md>), [teams](<https://devfeed.tech/tags/teams.md>)

### AI overview

All nine prize-winning teams in the first NEDO Quantum Challenge season used IBM quantum computers during research, and six ran their research exclusively on IBM systems. Projects covered infrastructure safety, trusted transactions, vehicle design, music and animation, low-carbon materials, drug discovery, and microbiome applications. IBM provided access to five quantum computers, education, programming sessions, consultations, and technical support; it also advised one team and co-researched with another. The article distinguishes research use from final results: Shionogi tested its drug-discovery approach on IBM hardware but produced final results by simulation, while other teams used IBM alone or alongside other providers. The excerpt also describes hybrid quantum-classical methods and named tools used by individual teams, without establishing that these approaches demonstrated practical quantum advantage.

### Source excerpt

Teams across all three tracks of Japan's national quantum competition used IBM quantum computers during research on real-world applications.

## Qiskit SDK v2.6 release summary

DevFeed: [Qiskit SDK v2.6 release summary](<https://devfeed.tech/articles/qiskit-sdk-v2-6-release-summary-96002.md>)

Original publisher: [Read original article](<https://research.ibm.com/blog/qiskit-2-6-release-summary>)

Author: Qiskit Team

Published: 2026-10-22T04:00:00Z

Content type: release

Language: en

Sources: [IBM Research](<https://devfeed.tech/sources/ibm-research.md>)

Topics: [SDK](<https://devfeed.tech/topics/sdk.md>), [Release notes](<https://devfeed.tech/topics/release-notes.md>), [version](<https://devfeed.tech/topics/version.md>)

Tags: [api](<https://devfeed.tech/tags/api.md>), [features](<https://devfeed.tech/tags/features.md>), [new-features](<https://devfeed.tech/tags/new-features.md>), [qiskit](<https://devfeed.tech/tags/qiskit.md>), [quantum-software-updates](<https://devfeed.tech/tags/quantum-software-updates.md>), [release](<https://devfeed.tech/tags/release.md>), [release-summary](<https://devfeed.tech/tags/release-summary.md>), [sdk](<https://devfeed.tech/tags/sdk.md>), [summary](<https://devfeed.tech/tags/summary.md>), [technical](<https://devfeed.tech/tags/technical.md>), [top](<https://devfeed.tech/tags/top.md>), [updates](<https://devfeed.tech/tags/updates.md>), [v2](<https://devfeed.tech/tags/v2.md>)

### AI overview

Qiskit SDK v2.6 expands the C API with a customizable pass manager, QPY serialization, and user-defined operations, giving compiled-language programs more control over circuit compilation and exchange without a Python round-trip. Passes can declare input and output intermediate representations so pipelines can be type-checked across multiple stages. C API QPY reads and writes native circuits compatible with Python's qiskit.qpy; format v18 is now the default and may reduce files with large or numerous registers, but older Qiskit releases may not load v18 files. The release also advances work toward standalone C API use without Python installed, though that support is not yet complete. Python 3.10 remains supported in v2.6 and is dropped in 2.7; source builds now require Rust 1.89. The C API is not yet declared stable, but this release's changes are additive.

### Source excerpt

Technical release summary for Qiskit SDK v2.6, including updates on top new features.

## A new plugin can enable fast runtime monitoring and control

DevFeed: [A new plugin can enable fast runtime monitoring and control](<https://devfeed.tech/articles/a-new-plugin-can-enable-fast-runtime-monitoring-and-control-92680.md>)

Original publisher: [Read original article](<https://research.ibm.com/blog/fast-interpretability-plugin-vllm>)

Author: Kim Martineau

Published: 2026-10-21T12:00:00Z

Content type: news

Language: en

Sources: [IBM Research](<https://devfeed.tech/sources/ibm-research.md>)

Topics: [vllm](<https://devfeed.tech/topics/vllm.md>), [AI Infrastructure](<https://devfeed.tech/topics/ai-infrastructure.md>), [CUDA Graphs](<https://devfeed.tech/topics/cuda-graphs.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [ai-transparency](<https://devfeed.tech/tags/ai-transparency.md>), [control](<https://devfeed.tech/tags/control.md>), [developers](<https://devfeed.tech/tags/developers.md>), [distributed-inference](<https://devfeed.tech/tags/distributed-inference.md>), [generative-ai](<https://devfeed.tech/tags/generative-ai.md>), [news](<https://devfeed.tech/tags/news.md>), [plugin](<https://devfeed.tech/tags/plugin.md>), [runtime-monitoring](<https://devfeed.tech/tags/runtime-monitoring.md>), [trustworthy-generation](<https://devfeed.tech/tags/trustworthy-generation.md>), [vllm](<https://devfeed.tech/tags/vllm.md>)

### AI overview

IBM's vLLM-Hook MIA lets developers capture or steer selected model internals during vLLM inference while retaining graph capture and replay. GPU operators set up routing entries before graph recording; during replay, those entries guide access to model states, and a courier transfers captured data to CPU memory. This approach avoids the eager execution mode required by many Python-hook tools, which disables graph replay. IBM reports that MIA was up to 20 times faster than its predecessor and up to 14 times faster than a similar code-insertion method. In reported steering experiments, overhead was 2% for an 8B Llama and Qwen 1.5B model in some cases, and 6% for Llama 70B. The excerpt describes potential uses such as flagging malicious prompts or adjusting responses, but does not establish performance beyond the reported experiments.

### Source excerpt

IBM's vLLM-Hook MIA lets developers inspect and edit the AI generation process with hardly any delay.

## llm-d 1.0 hardens open-source AI inference for production

DevFeed: [llm-d 1.0 hardens open-source AI inference for production](<https://devfeed.tech/articles/llm-d-1-0-hardens-open-source-ai-inference-for-production-91207.md>)

Original publisher: [Read original article](<https://research.ibm.com/blog/llmd-1-0-is-ready-for-production-inference-serving>)

Author: Peter Hess

Published: 2026-10-20T12:00:00Z

Content type: release

Language: en

Sources: [IBM Research](<https://devfeed.tech/sources/ibm-research.md>)

Topics: [LLM optimization](<https://devfeed.tech/topics/llm-optimization.md>), [AI Infrastructure](<https://devfeed.tech/topics/ai-infrastructure.md>), [Production Engineering](<https://devfeed.tech/topics/production-engineering.md>), [sglang](<https://devfeed.tech/topics/sglang.md>), [vllm](<https://devfeed.tech/topics/vllm.md>)

Tags: [accelerators](<https://devfeed.tech/tags/accelerators.md>), [agent](<https://devfeed.tech/tags/agent.md>), [agentic](<https://devfeed.tech/tags/agentic.md>), [agentic-coding](<https://devfeed.tech/tags/agentic-coding.md>), [agentic-systems](<https://devfeed.tech/tags/agentic-systems.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-for-code](<https://devfeed.tech/tags/ai-for-code.md>), [ai-inference](<https://devfeed.tech/tags/ai-inference.md>), [ai-infrastructure](<https://devfeed.tech/tags/ai-infrastructure.md>), [ai-investment](<https://devfeed.tech/tags/ai-investment.md>), [ai-roi](<https://devfeed.tech/tags/ai-roi.md>), [api](<https://devfeed.tech/tags/api.md>), [cache](<https://devfeed.tech/tags/cache.md>), [distributed](<https://devfeed.tech/tags/distributed.md>), [distributed-inference](<https://devfeed.tech/tags/distributed-inference.md>), [generative-ai](<https://devfeed.tech/tags/generative-ai.md>), [inference](<https://devfeed.tech/tags/inference.md>), [infrastructure](<https://devfeed.tech/tags/infrastructure.md>), [llm-d](<https://devfeed.tech/tags/llm-d.md>), [open-source](<https://devfeed.tech/tags/open-source.md>), [open-source-ai](<https://devfeed.tech/tags/open-source-ai.md>), [production](<https://devfeed.tech/tags/production.md>), [release](<https://devfeed.tech/tags/release.md>)

### AI overview

llm-d 1.0 integrates distributed inference techniques with production operations for Kubernetes-based model serving. Its features include multi-tenant flow control and prioritization, zero-downtime upgrades, and routing that can keep repeated requests near warm prefix caches; KV-cache offloading can preserve reusable context across GPU, CPU, and storage. The project works with inference engines such as vLLM and SGLang. In a reported agentic coding test, its prefix-aware router captured 97% of available cache reuse. A separate IBM and Red Hat demonstration ran 3,000 software-engineering agents on self-hosted infrastructure at 5 to 10 times less than the actual price of serving the same open model as a service. These are reported examples, not guarantees of savings for other deployments.

### Source excerpt

The distributed inference framework makes serving large models on your own infrastructure reliable and affordable for the agentic era.

## New course teaches how AI models run on modern accelerators

DevFeed: [New course teaches how AI models run on modern accelerators](<https://devfeed.tech/articles/new-course-teaches-how-ai-models-run-on-modern-accelerators-90317.md>)

Original publisher: [Read original article](<https://research.ibm.com/blog/coursera-dataflow-accelerator-architecture-specialization>)

Author: Kaoutar El Maghraoui

Published: 2026-10-19T12:00:00Z

Content type: article

Language: en

Sources: [IBM Research](<https://devfeed.tech/sources/ibm-research.md>)

Topics: [Learning](<https://devfeed.tech/topics/learning.md>), [NVIDIA Research](<https://devfeed.tech/topics/nvidia-research.md>), [coreboot](<https://devfeed.tech/topics/coreboot.md>)

Tags: [accelerators](<https://devfeed.tech/tags/accelerators.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-hardware](<https://devfeed.tech/tags/ai-hardware.md>), [ai-models](<https://devfeed.tech/tags/ai-models.md>), [ai-systems](<https://devfeed.tech/tags/ai-systems.md>), [algorithms](<https://devfeed.tech/tags/algorithms.md>), [c](<https://devfeed.tech/tags/c.md>), [c-plus-plus](<https://devfeed.tech/tags/c-plus-plus.md>), [cerebras](<https://devfeed.tech/tags/cerebras.md>), [columbia-university](<https://devfeed.tech/tags/columbia-university.md>), [computer-science](<https://devfeed.tech/tags/computer-science.md>), [generative-ai](<https://devfeed.tech/tags/generative-ai.md>), [ibm-research](<https://devfeed.tech/tags/ibm-research.md>), [labs](<https://devfeed.tech/tags/labs.md>), [research](<https://devfeed.tech/tags/research.md>), [specialization](<https://devfeed.tech/tags/specialization.md>), [technical-note](<https://devfeed.tech/tags/technical-note.md>)

### AI overview

IBM Research describes a six-course Coursera specialization that teaches engineers how AI models execute across five accelerator architectures: GPUs, Google TPUs, Cerebras WSE, Groq LPU, and IBM Spyre. The courses cover execution models, compilation, memory systems, debugging, and performance modeling, with hands-on labs using PyTorch. Core labs run on Google Colab; optional practice uses Spyre hardware.

### Source excerpt

'Dataflow Architectures for AI' is a 6-course specialization from IBM Research that compares leading accelerator architectures through hands-on labs.

## Preparing chip designers for the nanostack era

DevFeed: [Preparing chip designers for the nanostack era](<https://devfeed.tech/articles/preparing-chip-designers-for-the-nanostack-era-86268.md>)

Original publisher: [Read original article](<https://research.ibm.com/blog/preparing-chip-designers-for-the-nanostack-era>)

Author: Peter Hess

Published: 2026-10-13T16:00:00Z

Content type: news

Language: en

Sources: [IBM Research](<https://devfeed.tech/sources/ibm-research.md>)

Topics: [Chip design](<https://devfeed.tech/topics/chip-design.md>), [ESP32-C3](<https://devfeed.tech/topics/esp32-c3.md>), [3D](<https://devfeed.tech/topics/3d.md>), [Publishing](<https://devfeed.tech/topics/publishing.md>)

Tags: [advanced-logic](<https://devfeed.tech/tags/advanced-logic.md>), [chip-design](<https://devfeed.tech/tags/chip-design.md>), [design-technology-co-optimization](<https://devfeed.tech/tags/design-technology-co-optimization.md>), [eda](<https://devfeed.tech/tags/eda.md>), [hardware](<https://devfeed.tech/tags/hardware.md>), [hardware-technology](<https://devfeed.tech/tags/hardware-technology.md>), [logic-scaling](<https://devfeed.tech/tags/logic-scaling.md>), [nanosheet-transistors](<https://devfeed.tech/tags/nanosheet-transistors.md>), [release](<https://devfeed.tech/tags/release.md>), [semiconductors](<https://devfeed.tech/tags/semiconductors.md>), [silicon](<https://devfeed.tech/tags/silicon.md>)

### AI overview

IBM plans to release an early research and development process design kit for its nanostack chip technology in the first quarter of 2027. The kit will provide design resources for evaluating the architecture, which stacks transistor devices vertically, and is intended to help chip companies, EDA vendors, equipment suppliers, and researchers explore the technology and its design challenges.

### Source excerpt

IBM is releasing a process design kit for our groundbreaking 0.7-nanometer transistor architecture.

## Scaling AI-driven chip design with open agent factories

DevFeed: [Scaling AI-driven chip design with open agent factories](<https://devfeed.tech/articles/scaling-ai-driven-chip-design-with-open-agent-factories-86267.md>)

Original publisher: [Read original article](<https://research.ibm.com/blog/open-agent-factories-and-distributed-inference-for-chip-design>)

Author: Peter Hess

Published: 2026-10-13T12:00:00Z

Content type: article

Language: en

Sources: [IBM Research](<https://devfeed.tech/sources/ibm-research.md>)

Topics: [Chip design](<https://devfeed.tech/topics/chip-design.md>), [vllm](<https://devfeed.tech/topics/vllm.md>), [quantization](<https://devfeed.tech/topics/quantization.md>), [LLM security](<https://devfeed.tech/topics/llm-security.md>)

Tags: [advanced-logic](<https://devfeed.tech/tags/advanced-logic.md>), [agent](<https://devfeed.tech/tags/agent.md>), [agentic-ai](<https://devfeed.tech/tags/agentic-ai.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-hardware](<https://devfeed.tech/tags/ai-hardware.md>), [ai-systems](<https://devfeed.tech/tags/ai-systems.md>), [chip-design](<https://devfeed.tech/tags/chip-design.md>), [generative-ai](<https://devfeed.tech/tags/generative-ai.md>), [hardware](<https://devfeed.tech/tags/hardware.md>), [inference](<https://devfeed.tech/tags/inference.md>), [llm-d](<https://devfeed.tech/tags/llm-d.md>), [logic-scaling](<https://devfeed.tech/tags/logic-scaling.md>), [news](<https://devfeed.tech/tags/news.md>), [quantum-error-correction](<https://devfeed.tech/tags/quantum-error-correction.md>), [research](<https://devfeed.tech/tags/research.md>), [scaling](<https://devfeed.tech/tags/scaling.md>)

### AI overview

IBM Research describes using an agent factory and llm-d distributed inference to optimize a quantum error-correction decoder for hardware. In one 100-hypothesis search, the all-open-model setup achieved a 52x speedup in high-level synthesis estimates at about $26 in inference cost, compared with $299 for the all-proprietary setup, which achieved a 204x speedup. The article notes that these results come from individual search runs and are not a universal model ranking.

### Source excerpt

Open models and llm-d enabled IBM Research to explore quantum error correction hardware at 11.5x lower inference cost than an all-proprietary configuration.

## How distributed inference keeps large AI models moving

DevFeed: [How distributed inference keeps large AI models moving](<https://devfeed.tech/articles/how-distributed-inference-keeps-large-ai-models-moving-85494.md>)

Original publisher: [Read original article](<https://research.ibm.com/blog/distributed-inference>)

Author: Peter Hess

Published: 2026-10-12T04:00:00Z

Content type: article

Language: en

Sources: [IBM Research](<https://devfeed.tech/sources/ibm-research.md>)

Topics: [model-serving](<https://devfeed.tech/topics/model-serving.md>), [LLM optimization](<https://devfeed.tech/topics/llm-optimization.md>), [Distributed Systems & Parallel Computing](<https://devfeed.tech/topics/distributed-systems-parallel-computing.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [ai-for-code](<https://devfeed.tech/tags/ai-for-code.md>), [batch-processing](<https://devfeed.tech/tags/batch-processing.md>), [cache](<https://devfeed.tech/tags/cache.md>), [distributed-inference](<https://devfeed.tech/tags/distributed-inference.md>), [explainer](<https://devfeed.tech/tags/explainer.md>), [generative-ai](<https://devfeed.tech/tags/generative-ai.md>), [inference](<https://devfeed.tech/tags/inference.md>), [open-source](<https://devfeed.tech/tags/open-source.md>)

### AI overview

The article explains how distributed inference coordinates model servers across a cluster using llm-d. It covers prefix-cache-aware routing, KV-cache offloading, flow control, batch inference, prefill/decode disaggregation, and wide expert parallelism. These techniques aim to reduce repeated computation, manage competing workloads, and improve accelerator utilization as agent traffic introduces long contexts and uneven request bursts.

### Source excerpt

Distributed inference serving is an air traffic control tower for AI traffic, and agentic workflows make it more necessary than ever.

## What's new at IBM Quantum Q3 2026

DevFeed: [What's new at IBM Quantum Q3 2026](<https://devfeed.tech/articles/what-s-new-at-ibm-quantum-q3-2026-77395.md>)

Original publisher: [Read original article](<https://research.ibm.com/blog/whats-new-q3-2026>)

Published: 2026-10-07T04:00:00Z

Content type: release

Language: en

Sources: [IBM Research](<https://devfeed.tech/sources/ibm-research.md>)

Topics: [Quantum Computing](<https://devfeed.tech/topics/quantum-computing.md>), [ibm](<https://devfeed.tech/topics/ibm.md>)

Tags: [circuits](<https://devfeed.tech/tags/circuits.md>), [client](<https://devfeed.tech/tags/client.md>), [error](<https://devfeed.tech/tags/error.md>), [error-correction](<https://devfeed.tech/tags/error-correction.md>), [execution](<https://devfeed.tech/tags/execution.md>), [extend](<https://devfeed.tech/tags/extend.md>), [foundation](<https://devfeed.tech/tags/foundation.md>), [ibm](<https://devfeed.tech/tags/ibm.md>), [mitigation](<https://devfeed.tech/tags/mitigation.md>), [model](<https://devfeed.tech/tags/model.md>), [news](<https://devfeed.tech/tags/news.md>), [platform](<https://devfeed.tech/tags/platform.md>), [quantum](<https://devfeed.tech/tags/quantum.md>), [quantum-enablement](<https://devfeed.tech/tags/quantum-enablement.md>), [quantum-error-correction](<https://devfeed.tech/tags/quantum-error-correction.md>), [run](<https://devfeed.tech/tags/run.md>), [servers](<https://devfeed.tech/tags/servers.md>), [stack](<https://devfeed.tech/tags/stack.md>), [this](<https://devfeed.tech/tags/this.md>), [tools](<https://devfeed.tech/tags/tools.md>), [update](<https://devfeed.tech/tags/update.md>)

### AI overview

IBM Quantum has moved error mitigation that previously ran on its servers to the client, where users can see, customize, and extend it. The directed execution model underpins IBM's error correction stack and gives users control over how their circuits run. The excerpt does not provide details about other quarterly updates.

### Source excerpt

Explore the latest updates, releases, and resources from IBM Quantum and the Qiskit community.

## Qiskit beyond Python: Fortran, C++, and Julia

DevFeed: [Qiskit beyond Python: Fortran, C++, and Julia](<https://devfeed.tech/articles/qiskit-beyond-python-fortran-c-and-julia-65718.md>)

Original publisher: [Read original article](<https://research.ibm.com/blog/qiskit-beyond-python>)

Author: Haimeng Zhang; Jim Garrison; Matthew Treinish; Henry Zou; Aaryav Mishra; Dhinakaran Vinayagamurthy; Shobhit Pandey; Nate Earnest-Noble; Robert Davis

Published: 2026-10-06T04:00:00Z

Content type: tutorial

Language: en

Sources: [IBM Research](<https://devfeed.tech/sources/ibm-research.md>)

Topics: [High-Performance Computing](<https://devfeed.tech/topics/high-performance-computing.md>), [Fortran](<https://devfeed.tech/topics/fortran.md>), [c/c++](<https://devfeed.tech/topics/c-c-plus-plus.md>), [The Julia Language](<https://devfeed.tech/topics/julia.md>), [Python](<https://devfeed.tech/topics/python.md>)

Tags: [c-plus-plus](<https://devfeed.tech/tags/c-plus-plus.md>), [explainer](<https://devfeed.tech/tags/explainer.md>), [fortran](<https://devfeed.tech/tags/fortran.md>), [hpc](<https://devfeed.tech/tags/hpc.md>), [qiskit](<https://devfeed.tech/tags/qiskit.md>), [quantum-software](<https://devfeed.tech/tags/quantum-software.md>), [scientific-computing](<https://devfeed.tech/tags/scientific-computing.md>), [sdk](<https://devfeed.tech/tags/sdk.md>)

### AI overview

Qiskit's C API lets developers use its Rust core from Fortran, C++, and Julia without an intermediate Python layer. The article introduces the native bindings and demonstrates a Julia workflow for constructing, optimizing, running, and analyzing a quantum simulation, connecting Qiskit to existing HPC and scientific-computing code.

### Source excerpt

Qiskit's C API connects the SDK's Rust core to the languages of HPC and scientific computing, so you can build and run circuits natively--no Python required.

## Directed execution: same performance, more control

DevFeed: [Directed execution: same performance, more control](<https://devfeed.tech/articles/directed-execution-same-performance-more-control-65078.md>)

Original publisher: [Read original article](<https://research.ibm.com/blog/directed-execution>)

Author: Jessie Yu; Tushar Mittal; Robert Davis

Published: 2026-10-05T16:30:00Z

Content type: article

Language: en

Sources: [IBM Research](<https://devfeed.tech/sources/ibm-research.md>)

Topics: [High-Performance Computing](<https://devfeed.tech/topics/high-performance-computing.md>), [Distributed Systems & Parallel Computing](<https://devfeed.tech/topics/distributed-systems-parallel-computing.md>), [return](<https://devfeed.tech/topics/return.md>), [AI Infrastructure](<https://devfeed.tech/topics/ai-infrastructure.md>)

Tags: [ai-coding](<https://devfeed.tech/tags/ai-coding.md>), [ai-coding-assistant](<https://devfeed.tech/tags/ai-coding-assistant.md>), [deep-dive](<https://devfeed.tech/tags/deep-dive.md>), [hardware](<https://devfeed.tech/tags/hardware.md>), [ibm](<https://devfeed.tech/tags/ibm.md>), [quantum](<https://devfeed.tech/tags/quantum.md>), [quantum-error-correction-mitigation](<https://devfeed.tech/tags/quantum-error-correction-mitigation.md>), [quantum-hardware](<https://devfeed.tech/tags/quantum-hardware.md>), [quantum-software](<https://devfeed.tech/tags/quantum-software.md>)

### AI overview

IBM describes directed execution for Qiskit workloads, a model that exposes circuit transformations, randomization and error mitigation through client-side tools and the Executor primitive. The article explains how Sampler and Estimator use this framework, outlines workflows for combining mitigation methods, and discusses its use in quantum error-correction research on IBM hardware. It also describes an AI-assisted tool for migrating existing Qiskit Runtime code to the new client-side primitives.

### Source excerpt

See, customize, and extend how your circuits use the error mitigation and correction methods researchers are exploring today on IBM quantum hardware

[Next page](<https://devfeed.tech/sources/ibm-research.md?cursor=WyIyMDI2LTEwLTA1VDE2OjMwOjAwKzAwOjAwIiwgImZlMzZkODZkLTA1OTgtNGM1My1hN2I4LTQzNjRjZGY1MGViNSJd>)